The goal of this research to understand the molecular events that are required for high-level globin gene expression in erythroid cells. Therapy of sickle cell disease and the thalassemias, which has focused on increasing fetal globin expression, would be furthered by elucidating the mechanism of globin gene activation. Previous experiments have suggested that binding sites for Nuclear factor-erythroid 2 (NF-E2), in globin regulatory sequences, are important for globin gene expression (11,45,60,78). More recently, it has been shown that globin gene expression is not strictly dependent on NF-E2 (72). We present evidence that NF-E2 deficiency causes a partial block to erythroid maturation and globin gene expression. We also present evidence that proteins, other than NF-E2, activate globin expression through NF-E2 sites. It is our hypothesis that NF-E2 binding sites are required for globin gene expression, and that in the absence of NF-E2, other proteins are able to bind NF-E2 sites and activate globin gene expression. Therefore, the specific aims contained in this grant proposal are: 1) To determine the role of NF-E2 binding sites in globin gene expression; 2) To characterize the proteins that activate globin gene expression through NF-E2 sites; and 3) To gain insight into the mechanism of globin gene activation by NF-E2. To achieve each of these aims, a series of experiments is proposed in genetically modified mice and cell lines. To examine the role of NF-E2 sites in globin gene expression, ablation of tandem NF-E2 binding sites in the mouse alpha-globin positive regulatory element (alphaPRE) is proposed. Experiments are proposed to characterize changes in erythropoiesis due to loss of NF-E2. We propose to biochemically characterize proteins in extracts from primary erythroid cells. Studies are proposed, in transgenic mice, to identify proteins that activate globin gene expression in place of NF-E2. Structural information on a domain related to NF-E2, has recently been made available (69). We propose to functionally map this domain, as well as a second domain at the N-terminus of NF-E2. To identify molecules that interact with these motifs, we propose yeast one/two-hybrid screens. Results from these experiments should yield important insights into the regulation of globin gene expression.